Importing an Excel or Word file may fail with an error message mentioning tags
Posted by Péter Botta on 17 July 2013 12:22 PM
Title: Importing an Excel or Word file may fail with an error message mentioning tags
Description: With source files containing any "local" formatting, like a font size change, using a cascading filter may result in a failed import. The error message will be similar to this:
If you want to use cascading filters, like Excel and HTML
"chained" after it, then the first filter in the "chain"
must not produce any inline tags or memoQ tags. These tags are typically the
result of some local formatting, like a font change. The only real way to prevent
the creation of tags is to make sure there is no local formatting in the
document you are importing. (For example, in Excel, formatting apllied to a whole cell may be OK, but formatting applied to some words only may not.)
The explanation is that the filters you can use in a cascading filter
"chain" (like HTML) expect text as input. If you use a cascading filter, the output (import result) of the first filter is sent to the second filter. (And so on if there are further filters in the chain.). But the "normal"
input of those filters is a text file, and they cannot handle tags created by a previous filter.
If you want to see where a filter (for example, the Excel filter) creates a tag, import the same file normally, without a cascading filter (for example, directly with the Excel filter), and then observe the result, or use the filtering toolbar to filter for rows in which the source segment contains tags.
The regex tagger filter is an exception to this: you can chain the regex tagger after any other filter in any case, even if the filter before it produces tags.